[Nut-upsuser] Server Does Not Shut Down - SOLVED!

Steve Read sd_read at yahoo.ca
Mon Feb 27 22:25:31 UTC 2012

Thank you for your reply Charles.

First you are correct in both assumptions, my microprocessor does provide the voltages and my table listing Pin 9 should read Pin 8.

I spent some time reading and thinking about your email and realized that normal (AC present, battery normal) is at -12V which is a logic low.
I thought this was how I designed it but when I got funny results I convinced myself that it was +12V.

Moving on, I have never heard of statserial but I found it on the install CD and tried it out and it is a wonderful tool that I won't forget.

Setting the switches for -12V provided normal operation of logic 0 and flipping either pin 1 or pin 8 caused that signal to change to a 1. 

HOWEVER, flipping both to +12V caused both lines to show 0 again? 

The only possible way for this is if the ground pin 5 was broken. I moved the ground wire of my simple jig from pin 5 and connected it to a metal point on the server (which I figured was ground) and eureka! 

I exited out of statserial, started up nut and this time setting both lines to +12V caused the servers to shut down as expected.

So, pin 5 (ground) is broken somewhere which I confirmed with my meter so tonight I will have to disassemble this server and fix the ground issue.

Thank you for your help, I am eager to fix this and get things back to normal. 



 From: Charles Lepple <clepple at gmail.com>
To: Steve Read <sd_read at yahoo.ca> 
Cc: "nut-upsuser at lists.alioth.debian.org List" <nut-upsuser at lists.alioth.debian.org> 
Sent: Sunday, February 26, 2012 5:55 PMbrocken
Subject: Re: [Nut-upsuser] Server Does Not Shut Down
On Feb 26, 2012, at 3:14 PM, Steve Read wrote:

> I made a simple device to apply either +12 or -12volts to both pin 1(DCD) and pin 8 (CTS) to help me understand the correct handshake but I am very confused.
> Here is a table of what I get:
> State    Pin 1    Pin 9    Note         What I understand it means
> 1        +12V    +12V    OL            On Line - normal operation.
> 2        -12V    +12V    LB OL        I don't understand this. It should be On Battery.
> 3        +12V    -12V    OB            I thought this should be Low Battery.
> 4        -12V    -12V    OL            I thought the server should shut down?
> Note: results of running:
> /usr/local/ups/bin/upsc sdrups at localhost ups.status

For reference, here is the block for type=9 in genericups.h:

        /* Type 9 */
        { "APC",
          "APC Back-UPS (940-0023A cable)",
          0,                            /* cable power: none            */
          TIOCM_CD, 0,                  /* online: CD off               */
          TIOCM_CTS, TIOCM_CTS,         /* low battery: CTS on          */
          TIOCM_RTS                     /* shutdown: RTS                */

And the text from the genericups man page:

- - - - - - - - - -

9 = APC Back-UPS/Back-UPS Pro/Smart-UPS with 940-0023A cable

[CP=none] [OL=-DCD] [LB=CTS] [SD=RTS]

- - - - - - - - - -

For your tests where you apply voltage directly, CP is not relevant. (I assume your microcontroller provides its own power to the serial port control lines.) Also, SD=RTS is the shutdown signal *to* the UPS (saying that NUT acknowledges the UPS' low battery condition). But you mentioned pins 1 (DCD) and 8 (CTS), so we can ignore that for now.

The way I read this, online with a good battery should be negative voltages on both DCD and CTS. OB+LB should be positive voltages on both. On Battery (but not yet low) should be DCD high, CTS low. (The fourth condition could occur if the power has returned, but the battery has not yet charged.)

I am confused as to how you could get OL for both +/+ and -/- states. Your table says Pin 1 and Pin 9, but earlier you mention Pin 1 and Pin 8. Could pin 8 have been floating during the voltage test?

You can use statserial to double-check that your serial port hardware is still good.

Charles Lepple
clepple at gmail
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.alioth.debian.org/pipermail/nut-upsuser/attachments/20120227/999e8d49/attachment-0001.html>

More information about the Nut-upsuser mailing list